Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
Supply Chain Implications of Customer-Centric Merchandising
Strategy
Rahul Tyagi
Associate Partner, IBM
_______________________________________________
Key Words
Customer-centricity, merchandising, assortment planning, supply chain, space
planning, planogram, logistics, retail strategy
Abstract
With retailers looking to adopt a customer-centric strategy to succeed in the hyper-competitive
environment, it is vital to understand the impact of such a strategy on a retailer’s business
functions. While the impact on merchandising is well understood in terms of sharper assortment
and improved presentation, the impact on supply chain needs to be studied in detail. This would
ensure the alignment of customer-centricity with the organization’s overall direction and growth
of revenues and profitability while increasing customer satisfaction levels.
_____________________________________
Introduction
With the rapidly changing customer demographic, retailers have realized the need to
cater to the tastes and preferences of the segments of the population which were earlier
lost in the generalization of standard merchandising practices. In addition, because of
channel-blurring, the retail space has become so hyper-competitive that a retailer has to
differentiate in order to ensure longevity. Such differentiation strategies are manifesting
themselves either through the adoption of a value strategy where the focus is on giving
best value to the customer or through a services & offerings strategy where the focus is
more on providing value-added services & a better offering to the customer to enhance
the shopping experience. Most of the extant retail merchandising practices are partial
towards planning at a retail chain or a geographic division level ignoring the impact of
variations in demographics across stores within a division. Such strategies also tend to
diffuse the impact of the retailer’s professed go-to-market strategy leading to
incongruence between the strategic and the operational goal.
What customer-centric strategy entails
Customer-centric merchandising calls for the study of consumer buying trends at a
store-level and then grouping the stores based on similarity of consumer behaviour.
Thus, one would end up with a finite number of store clusters with each cluster
exhibiting some commonality in consumer buying patterns. The sales data of the stores
coupled with market basket analytics would also reflect the consumer buying trends.
1
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
Subsequent to obtaining clusters, the assortment and space planning process would be
carried out at cluster level. This would be a true customer-centric merchandising as the
planning processes become aligned to the customer’s preferences and tastes. However,
this raises the prospect of having as many different assortments as there are store
clusters. Thus, a retailer would move from a planning process where there was a
standard assortment across the entire chain or a division to one where there are as
many assortments as there are store clusters.
This process change has significant supply chain implications. It is likely that for
an individual store the number of items in its assortment may come down yet at the
same time the total number of items across all assortments may go up. It is also likely
that a distribution centre which was supplying a standard assortment of items to all
stores attached to it may have to supply different set of items to different stores
belonging to different store clusters and may have to carry a larger set of items.
The level of process change at a warehouse would be dictated by:
1.
The number of different clusters to which the stores it services belong to: The
higher this number, the larger will be the assortment of items which the
warehouse has to carry. Increased assortment for same or slightly higher
volumes is likely to increase operational costs.
2.
The configuration of the warehouse: A modern warehouse with dynamic slot
allocations capable of having flexible size bins can more easily accommodate a
larger assortment and a lower average volume per item.
One possibility may be to re-assign stores to distribution centres to ensure that
stores in the same cluster are serviced by the same distribution centre. However, this
may not always be feasible as clusters may have stores which are geographically
dispersed and re-alignment with a warehouse may incur higher transportation costs
and lead times. A more achievable goal is to minimize the number of clusters serviced
by a warehouse. This may call for a re-look at the retailer’s distribution strategy. It may
merit evaluating a multi-level distribution strategy where key warehouses service RDCs
(Regional Distribution Centres) which in turn service much smaller local DCs or store
fulfilment centres. An RDC can be serviced by more than one warehouse and a local DC
can be serviced by multiple RDCs.
Cluster-level assortment planning may also have an implication on how the
merchandise is moved to the store from the DC. A good practice is to classify items both
by their velocity out of DC and by the number of stores serviced by the warehouse that
carries that item. Table 1 is a case in point where transportation strategies are evaluated
in the light of ‘Velocity of an item through DC’ and ‘Number of stores that carry that
item’.
Cost-benefit analysis of a customer-centric merchandising strategy
It would be almost tautological to state that to best serve the customer, it is important to
tailor the store assortment to the needs and wants of the customer. However, in doing
so retailers face the trade-off between increased supply chain costs and the right level of
2
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
assortment customization. An approach is presented here which can help retailers and
CPG companies to better evaluate micro-merchandising decisions in the light of supply
chain implications. This will also help retailers identify the optimal level of micromerchandising.
At an overall level, the supply chain costs will start going up as the retailer
decides to move from a chain-wide planning to store-cluster level planning. However,
the increase will not be significant initially where the number of clusters is limited. As
the number of clusters approaches the number of stores there will be an inflection point
where the incremental benefits of micro-merchandising will be offset by increase in
supply chain costs required to support such a program. To better understand the costbenefit trade-off of micro-merchandising we will have to look at the various
components which define the cost-benefit equation.
Key benefit elements
Some of the key benefits arising out of micro-merchandising are:
1.
Increased sales: Since the assortment is more in line with the clientele visiting
the store, hence, the most likely outcome of such an initiative would be an
increase in the total sales for the store.
2.
Increased profits: A store with merchandise tailored to the customer needs is
likely to increase the satisfaction levels of the shoppers. Higher levels of
customer satisfaction can be used to drive better pricing by the retailer as the
customers would be willing to pay more in return for a better selection.
3.
Effective new product introductions: Introducing a new product today is like
shooting in the dark and the retailers have to assess early sales patterns to judge
where the products would do well and which stores to pull the products out of.
By matching the characteristics of store-clusters with the attributes of a new
product, retailers can choose the relevant clusters to launch the product in.
4.
Reduced lost sales & cost of carrying inventory: One of the problems of a using
a standardized assortment is that the store shelf will have relatively less space for
items which are sought after by the customers and relatively more space for less
wanted items. As a result, the store often ends up with empty shelves for the
former and little movement for the latter. This translates to lost sales for the
popular item and high inventory holding cost for the less desired items. Tailored
assortments do away with both these issues as not only are the items prioritized
based on the customer preferences but space allocation on shelf is also guided by
the principle of salability for that store or cluster.
5.
Alignment of pricing and promotions: Most retailers carry out pricing and
promotions for groups of stores. If this grouping can be aligned to the storeclusters which are obtained during the process of customer-centric
merchandising, then it is possible to tie the processes of assortment planning,
pricing and promotions together. Since the basis of micro-merchandising is
similarity in consumer behavior, the same basis would serve as an excellent
3
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
6.
proxy for price & promotions groups as well. After accounting for logistical
impacts of carrying out assortment planning, pricing and promotions planning
for the same store-cluster one can align these processes leading to savings in
labor costs and operational synergies for the merchandisers and category
managers.
Reduced Markdown and promotional expenses: A customer-driven
merchandising strategy would ensure higher sell-through. This would hold very
true in case of seasonal & fungible merchandise where the value of inventory
goes down with time. A better understanding of the store-cluster level customer
would enable a better match between the products and the clientele leading to
less end-of-season merchandise which effectively translates into lower
markdown costs associates with liquidating excess stocks.
Key Cost Elements
Having outlined the merits of a customer-centric strategy, one needs to delve into the
costs associated with such a strategy to enable an accurate estimation of the net benefit
that may accrue if a retailer chooses to deploy the aforementioned strategy. The costs
can be summarized under a few broad headings like:
a.
Process costs: Moving from a standardized assortment for a chain to a cluster
level assortment plan would entail increased demand on the time of the
merchandisers and category managers. One option is to automate the process by
correlating product attributes with the cluster’s demographic attributes to enable
a better product prioritization. This would enable the retailer to identify the key
product attributes which are important for the customers in that store-cluster. If
that is achievable then the business user participation can be limited to
assortment review and the formulation of business rules based upon
merchandising objectives for the cluster. The latter would be a one-time exercise
with periodic revisits to amend or append the rule set based upon changes in the
business clime in the intervening period.
b.
Supply chain costs: This more measurable and visible part of the cost increase
would arise out of planning for the replenishment for a different set of items for
different store-clusters. The execution would also entail cost increases owing to
impact on warehousing and logistics processes.
To get a better grasp on the impact on supply chain costs, it is imperative that one tries
to understand the various components of supply chain processes that are impacted.
Below are some of the key areas which will see a change in cost structure owing to the
adoption of a customer-centric merchandising strategy:
1.
Planning & ordering costs: The cluster assortments have a limited amount of
assortment overlap and the summation of various cluster-level assortments would
increase the total number of items for the chain. As a result
a.
The warehouse planning and ordering costs will increase: Not only will
warehouses have to plan for additional items but there is likely to be an
increase in the number of orders leading to a cost increase. This is so because
a warehouse may span multiple store-clusters, hence, multiple assortments
with a broader range of item velocity.
4
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
b.
Store’s planning & ordering costs will decrease: Stores will have to plan for
fewer items and their ordering frequency will decrease leading to a lower
ordering cost. The latter will be a result of aligning shelf space with customer
preferences leading to more space being allocated to popular items and
therefore infrequent stock-outs of those items and longer shelf replenishment
cycles.
As the stores’ planning and ordering costs are significantly higher than the warehouses’
costs for planning & ordering, hence, this would lead to a cost savings as the same
degree of impact yields a larger quantum of cost change at store level as compared to a
warehouse.
2.
Store operations cost: Stores will have fewer items to deal with. Shelf stocking
costs are likely to come down as more space is allocated to popular items leading
to lower replenishment frequency. Labor costs can also be saved if item shelving
rules like “case and a quarter” or case-plus can be adopted to enable case shelving.
This would do away with the costs associated with breaking a case before
shelving.
3.
Inventory costs: There would be profound impact on inventory costs. Warehouse
will now have to stock many more SKUs with a broader range of item velocity as
compared to earlier and not all SKUs will be supplied to all stores. This will
increase the inventory holding costs at the warehouse. Stores will see the positive
side of inventory impact as many slower items are eliminated leading not only to
freeing up valuable shelf space but also reducing holding costs. Owing to the high
velocity of the remaining items in the assortment and the reduction in the average
time an item spends on the shelf, there is likely to be a positive margin impact.
Overall, the inventory costs are likely to come down as the positive cost impact at
the store offsets the cost increase at the warehouse.
4.
Warehousing & logistics cost: Increase in the number of items to be stocked at the
warehouse entails an increase in the number of stocking & picking area
assignments. Space will have to be carved out for the increased number of SKUs
and since many of these items may have less movement, allocating bins will
become an issue. The shelving costs will also increase with an increase in the
number of bins and shelving areas in the warehouse. One standard best practice
followed by many retailers is to divide the warehouse into different zones
depending upon the type of items. Under such an arrangement, a “flex-area” is
dedicated to seasonal and low-volume/high-velocity items. Picking in such areas
may require specialized equipment like pick-to-light, robotic picking systems and
automated sortation equipment which enables less-than-case picks. Overall,
picking costs are likely to increase as pick routes become longer. Flow path
redesign can be resorted to in extreme conditions if picking costs increase beyond
expectations. Modern warehousing practices like AS/RS (Automated Storage &
Retrieval System), dynamic slotting and interleaved picking can substantially
mitigate the impact of SKU proliferation. The latter two practices have been found
especially effective in cases where the number of items is large and pick volumes
vary widely across products and time.
5
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
Logistics costs in terms of transportation costs are likely to go up as well. The
increase in costs will be felt on both the inbound and outbound sides of the
warehouse. Inbound costs will increase because there will be many SKUs with
smaller volumes (owing to limited number of stores carrying them) which will
require LTL or shipment consolidations. Both involve costs in terms of money or
time. Handling increased number of SKUs at warehouse inbound will increase
receiving costs and QC costs as well. On the outbound side, there will be two
countering influences. For each store a potential reduction in the number of overall
items and slow movers would increase shipment efficiency. More pallet level &
case level shipments can be planned leading to reduction in handling and shipping
costs. However, owing to the variation in the assortment across stores, route
planning will become complicated. Route plans will have to be generated at
multiple levels considering factors like common set of items, unique items per
store in stores on a common route, their replenishment frequency, loading
restrictions and consolidation opportunities. This is likely to lead to a polarization
in shipment strategies to stores. Unlike earlier when most shipments were mixed
and multi-store, in the new scheme of things there will be more dedicated
shipments to a store and fewer across store shipments. Of course, this has to be
viewed keeping in mind the replenishment frequencies of the common items
carried by stores along a common route and those of the unique items. At an
overall level, though, the outbound costs are most likely to go up.
a.
DSD item supply chain: One key pre-requisite for stores that are being
confined to a cluster is that they should be capable of receiving the items that
are assigned to them by the store-cluster level assortment. This issue is
exacerbated for DSD items as most DSD vendors have limited geographical
reach and if the cluster is geographically dispersed, then one would need to
create source of supply for the DSD items. This is a reason why lot of retailers
look at supply constraints before embarking on a customer-centric
merchandising initiative.
There are many qualitative factors that need to be taken into consideration. If the
item is a key item in the assortment, then the retailer would want to take control of the
supply chain of the item. On the other hand, if the item is a high-cost item with
infrequent sales, then it is best handled by the DSD supplier as that would remove the
logistics hassles from retailer’s consideration. Perishability is an important determinant
of making decision about going DSD for an item. Since freshness is of paramount
importance in such cases, one would want the item to be directly supplied to the store
rather than spending a part of replenishment lead time in the warehouse. Moving to
DSD is likely to increase the costs for an item as the vendor will take on the logistics
responsibility from the retailer and would expect to be defrayed for this additional
service. This premium is usually built into DSD pricing. Thus, one can see that there are
many supply chain functions which are accretive to the benefits of a customer-centric
merchandising strategy and there are certain which add to the costs of such an
initiative. Please refer Appendix for more details.
Strategic Factors in Decision-Making
6
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
In addition, retailers need to evaluate their customer-centric merchandising decisions in
the light of two major factors:

Their positioning in the market: Value players will have difficulty in raising
prices to absorb the cost impact whereas service-oriented retailer may be able to
play the market-focused assortment card quite well and counter any cost
increases with a perceived better offering. A customer-centric strategy may
reduce dependence on promotions facilitating a transition from a Hi-Lo pricing
to a more EDLP (Every Day Low Pricing) strategy. An EDLP strategy greatly
simplifies the pricing structure and the associated costs and is becoming
increasingly popular in the grocery retail industry.

Supply chain strategy: It may merit evaluating outsourcing options in
warehousing and transportation on a selective basis. The basis can be a choice of
products and stores which are leading to large increase in supply chain costs.
Simulation can be used to determine average supply chain cost per item-store in
a cluster. Items & stores that seem to be having costs consistently above the
average can then be evaluated for alternate fulfillment models. Some of the
commonly employed models would be:
 3PL: Retailer may want to look separately at warehousing and
transportation costs and then decide which portions of the supply chain to
outsource. There are specialized LTL service providers & supplier
shipment consolidators who can take over the required functions and
easily integrate with the retailer’s supply chain
 DSD: Always a good option if there are local suppliers available for the
item-store combination. Cost comparison needs to be done with other
supply options before deciding to go with DSD.
 Supply from another store: A good practice if the retailer has a large store
with a big backroom and when the cost of holding inventory in the store’s
backroom is not that high. This would just add another level to the
transportation network and would optimize the load factor as multiple
smaller runs are possible between stores with a limited cost increase.
One can derive a heuristics which can help in formulating the micromerchandising strategy from a supply chain point-of-view. This is illustrated as a
toolkit at the end of this article. In conclusion, it is imperative in the current retailing
environment for retailers to formulate a well-defined customer-centric merchandising
strategy. Yet in the course of operationalizing such a strategy one needs to look at the
impact it may have on various supply chain elements like warehousing, transportation,
replenishment and store operations. It is then that the true cost-benefit of a micromerchandising strategy can be explored and the right level of micro-merchandising
ascertained. Without this due diligence a retailer may end up increasing the overall
operational costs of such an initiative which may have repercussions on the bottom-line.
For retail segments like grocery which survive on wafer-thin margins this impact can be
fatal. There are plenty of examples in the industry where a half-baked customer-centric
merchandising strategy was put in place only to be hastily recanted owing to the
disproportionate rise in supply chain costs leading to bleeding of overall margins.
7
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
However, a well-conceived customer-focused merchandising strategy which marries
the benefits of increased customer-led assortment with the associated costs of achieving
that would not only ensure a higher level of customer satisfaction but would also lead
to a veritable increase in margins and cost containment.
Appendix: Micro-Merchandising Toolkit
1. Identify the basis for customer-centric merchandising. Since the general notion
is to make assortment more customer-centric, one would attempt to cluster stores
on the basis of factors that indicate customer behavior. These can be
demographics & geography related factors. POS information can, by reflection
(assuming sales represent customer interests), also be used to group stores.
2. Identify any constraints for the customer-centric store grouping process. These
constraints can be source of supply driven or regulatory (E.g. liquor, tobacco)
3. Evaluate the cost-benefit equation for the most granular level of store
clustering. This can be done using simulation with the various cost-benefit
elements as outlined below.
General Factors
Impact Factor
Sales
Gross Profit
New Product
Introduction
Costs
Lost Sales
Type of Impact
Increase
Notes
Shelf Item Sales(new state) - Shelf Item
Sales(old)
Shelf item sales under the new state can
be taken as (total shelf area) * (popular
item sales-old state) / (area allocated to
popular items-old state)
Increase
Profit from shelf(new) – Profit from shelf
(Average item (old)
sales
velocity Average margin of shelf = ∑(Item sales
increases,
velocity)*(Item margin)
average margin
may increase)
Decrease (The New products launched only in store
chances
of clusters where the probability of success
success will go is high. This can be achieved by
up)
matching the characteristics of a storecluster with item attributes and
identifying which are the key purchase
determinant item attributes.
Decrease
Popular items get more shelf space
owing to elimination of slow movers. A
good proxy for lost sales reduction will
be:
Average Inventory on Shelf of popular
items / Sales velocity of popular items
This would go up in customer-led
merchandising which implies less lost
sales
8
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
Supply Chain Factors
Impact Factor
Type of Impact
Ordering/Planni Decrease (Store
ng Costs
cost
decrease
would
more
than
offset
warehouse cost
increase)
Notes
Ordering costs are proportional to:
Line items in order * Order frequency
Warehouse: Increase in number of items
and most likely order frequency implies
cost increase.
Store: Reduction in number of items and
order frequency implies cost increase
Store Operations Decrease
More shelf space to popular items means
Cost
that re-stocking frequency decreases.
Also, full case packs can be stocked on
shelf rather than part-cases. Both imply
labor cost savings.
Inventory Costs
Decreases. Store Warehouse: Increase in item variety and
cost
decrease increase in complexity of replenishment
offsets
& fulfillment frequencies implies cost
warehouse cost increase.
increase
Store: Reduction in slow mover implies
cost savings. A good proxy for inventory
savings is:
∆ (Average Sales velocity of shelf) * Cost
Complement i.e. average factor to covert
retail price to cost. Sales velocity
increases which implies inventory cost
decrease.
Warehousing & Increase
Warehouse: Stocking & picking costs
Logistics Costs
increase and so do receiving costs.
Logistics: Cost per mile per unit
weight/volume goes up as route
planning becomes complex for both
inbound and outbound and LTL
increases.
To get a quantitative value for the
change in costs supply chain simulation
will have to be done. This can be
achieved using popular tools like eSCOR or simulation languages like
ARENA.
4. After one has computed the net cost-benefit impact at the most granular level of
customer-centric merchandising which is likely to be store-specific assortments,
one need to start clustering stores based on the chosen basis of micromerchandising.
5. Compute the cost-benefit equation for each level of store clustering.
6. Ideally, one should be able to come up with a graph of net benefit and level of
store clustering. (Figure 1) That graph can be used to identify the right level of
9
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
store clustering which aligns with the retailer’s customer-centric merchandising
strategy.
10
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
Tables & Graphs
Table 1: How to decide logistics strategy for an item
High


Velocity
Through
DC



More frequent shipments
to the “carrying” stores;
Impact area: Route
Planning
No significant impact on
shipping costs


Higher shipment frequency
Opportunity for load
consolidations (LTLs to
TLs); Shipping cost savings
possible
LTL/parcel shipping is the
plausible option
Shipping costs likely to
increase
Good candidates for DSD
(Direct to Store Delivery)

Evaluate load consolidation
opportunities to convert LTL
to TL
Shipping costs likely to
increase
DSD can be a viable option
(needs to be evaluated)


Low
Low
Number of stores
carrying that item
11
High
Journal of Business and Retail Management Research (JBRMR) Vol. 2 Issue 2 April 2008
Figure 1: Value Equation of Customer-Centric Merchandising
Net
Benefit
StoreSpecific
Level of Store
Clustering
12
ChainWide